Professional Documents
Culture Documents
Carlos Kiffer1, Andre Hsiung1, Carmen Oplustil1, Fleury Institute, Advisory Group on Antimicrobials and
Jorge Sampaio1, Elsa Sakagami1, Philip Turner2, Clinical Microbiology1, São Paulo, Brazil; AstraZeneca
Caio Mendes1, and the MYSTIC Brazil Group Pharmaceuticals2, Macclesfield, Cheshire, England
Establish the susceptibility pattern of Gram-negative bacteria causing infections in ICU patients,
MYSTIC Program Brazil 2003. Gram-negative bacteria (n = 1,550) causing nosocomial infections
were collected at 20 Brazilian centers. The central laboratory confirmed the identification and
performed the susceptibility tests by Etest methodology (AB Biodisk, Solna, Sweden) for
meropenem, imipenem, ciprofloxacin, ceftazidime, cefepime, cefotaxime, piperacillin/tazobactam,
gentamicin, and tobramycin. Interpretation criteria used were according to National Committee
for Clinical Laboratory Standards (NCCLS). Pseudomonas aeruginosa (30.3%) was the most
frequent isolate, followed by E. coli (18.6%), Klebsiella pneumoniae (16.9%), Acitenobacter
baumannii (8.8%), and Enterobacter cloacae (7.1%). Pseudomonas aeruginosa (n=470) isolates
presented susceptibility rates of 64% to meropenem, 63.8% to piperacillin/tazobactam, 63.4%
to amikacin, 58.7% to imipenem. Acitenobacter baumannii presented susceptibility rates to
meropenem of 97.1%, and 73% to tobramycin. E. coli and K. pneumoniae were highly susceptible
to both carbapenems.Carbapenem resistance among the Enterobacteriaceae is still rare in the
region. Acitenobacter baumannii and P. aeruginosa presented elevated resistance rates to all
antimicrobials. Since they play an important role in nosocomial infections in this environment,
the use of empirical combination therapy to treat these pathogens may be justified.
Key Words: Drug resistance, bacterial, microbial sensitivity tests, infection control, carbapenems.
A major issue confronting organized health care today changes in the spectrum of microbial pathogens and
is that of controlling the increase in antimicrobial trends in antimicrobial resistance patterns in nosocomial
resistance [1-4]. Although multiple factors play a role and community-acquired infections along time. The
in this problem, the selective pressures induced by information generated by surveillance programs,
inappropriate and widespread use of antibiotics are associated with an increased awareness about evolving
considered important contributors. Several studies have resistance patterns, have proved helpful for the
reported higher rates of antimicrobial resistance among development of empirical approaches for the treatment
isolates from intensive care units (ICUs) than among of serious infections [8]. Additionally, surveillance
isolates from general-patient-care areas [1,5-7]. These programs may also be useful in the prevention and
studies have provided important information about control of infections caused by resistant organisms [3-
5,7-13]. Furthermore, surveillance programs have
Received on 10 January 2005; revised 07 June 2005. provided evidence of important differences in
Address for correspondence: Dr. Carlos R.V. Kiffer. Fleury antimicrobial resistance patterns occurring in various
Institute/Advisory Group on Antimicrobials and Clinical
Microbiology. Avenida General Waldomiro de Lima 508, Zip geographical areas and even units within a certain area.
code: 04344-070 – Jabaquara – São Paulo – SP – Brazil.Phone/ However, those programs have a limited ability to
fax: +55 11 5014 7601. E-mail: carlos.kiffer@fleury.com.br identify and analyze all the relevant risk factors
associated with the different resistance patterns.
The Brazilian Journal of Infectious Diseases 2005;9(3):216-224 The Meropenem Yearly Susceptibility Test
© 2005 by The Brazilian Journal of Infectious Diseases and
Contexto Publishing. All rights reserved. Information Collection (MYSTIC) is a global, annual
www.bjid.com.br
BJID 2005; 9 (June) The MYSTIC Program Brazil 2003 217
and multicenter surveillance program that compares the to the investigators) were randomly selected for
activity of several broad-spectrum antimicrobial agents inclusion in this study. Multiple isolates of the same
in carbapenem user centers. MYSTIC Program Brazil species from a single origin (same patient) were
was started in 1999, involving three centers (ICUs excluded. Catheter, tracheal aspirates and
only); it was increased to seven centers in 2001 and bronchoalveolar lavage samples were submitted to
2002 (ICUs only), and matured to the present 2003 semi-quantitative/quantitative cultures, accordingly.
edition, with 20 centers (12 ICUS, 2 neutropenic patient Each participating laboratory performed identification
units, and 6 general wards). of microorganisms. The central laboratory (Fleury
The objective of our study was to determine the Diagnostics) confirmed the identification through
susceptibility pattern of Gram-negative bacteria causing conventional biochemical methodology or through the
nosocomial infections in hospital patients, as part of Vitek automated system.
the fourth edition of MYSTIC Program Brazil during
2003. It is our intention that these data could then be Susceptibility Tests
used locally, in conjunction with other related studies,
to properly interpret significant resistance patterns and The central laboratory determined the minimum
choose the most appropriate antimicrobial regimens for inhibitory concentrations (MICs) of meropenem,
empirical therapy. imipenem, ciprofloxacin, ceftazidime, cefepime,
cefotaxime, piperacillin/tazobactam, gentamicin,
tobramycin, and amikacin by Etest methodology (AB
Material and Methods Biodisk, Solna, Sweden) and interpretations were made
according to National Committee for Clinical
Details of the study design and susceptibility testing Laboratory Standards (NCCLS) [16]. Control strains
methods have been previously described [14,15]. of E. coli (ATCC 25922), E. coli (ATCC 35218),
and Pseudomonas aeruginosa (ATCC 27853) were
Participating Centers tested with each set of MIC determinations.
There were 19 participating centers during the 2003 Screening for Extended Spectrum β-Lactamase
program edition. All centers were asked to submit up (ESBL)
to 100 Gram-negative bacteria samples, representative
of the infectious process, regardless of the sample E. coli and K. pneumoniae with MICs ≥ 2 µg/mL
source, from specialized hospital units. All isolates were to any cephalosporins were submitted to an ESBL
collected from January to October 2003 from production test by double-disk synergy with amoxicillin/
hospitalized patients in 12 ICUs, 2 neutropenic patient clavulanic acid and ceftazidime, ceftriaxone,
units, and 6 general wards. Among the participating cefotaxime, and aztreonam. Isolates with an enhanced
centers, 10 were located in southeastern (7 in São zone for any of these agents and amoxicillin/clavulanic
Paulo, 2 in Rio de Janeiro, 1 in Minas Gerais), 7 in acid were considered ESBL producers for the purpose
southern (4 in Rio Grande do Sul, 2 in Paraná, 1 in of this report [17], since this test is not recommended
Santa Catarina), 2 in northeastern (Bahia), and 1 in by the NCCLS for confirmation of ESBL production.
midwest Brazilian states (Brasilia) (Table 1). Control strains K. pneumoniae (ATCC 700603 –
ESBL positive) and E. coli (ATCC 25922 – ESBL
Isolates negative) were assayed with each test set.
For the remaining species of Enterobacteriaceae,
One thousand five hundred and fifty Gram-negative isolates that produced intermediate to resistant MICs
bacilli responsible for the infectious process (according to cefepime were interpreted as compatible with a
www.bjid.com.br
218 The MYSTIC Program Brazil 2003 BJID 2005; 9 (June)
Table 1. Number of isolates (n) and contribution (%) Table 2. Prevalence of microorganisms isolated
per center – MYSTIC Program Brazil 2003
Microorganism N %
Center N % P. aeruginosa 470 30.3
1 97 6.3 E. coli 288 18.6
2 55 3.5 K. pneumoniae 262 16.9
4 83 5.3 A. baumannii 137 8.8
5 65 4.2 E. cloacae 110 7.1
6 59 3.8 P. mirabilis 51 3.3
7 86 5.5 S. maltophilia 42 2.7
8 74 4.8 S. marcescens 38 2.5
9 100 6.5 E. aerogenes 23 1.5
10 43 2.8 C. freundii 21 1.3
11 100 6.5 K. oxytoca 12 0.8
12 77 5.0 M. morganii 12 0.8
13 116 7.5 Others 84 5.4
14 96 6.2 Total 1,550 100
15 53 3.4
16 118 7.6
17 42 2.7
18 100 6.5
19 98 6.3
20 35 2.3
Total 1,550 100
N (%)
Microorganism Blood/Catheter Respiratory tract Urinary tract Skin/Soft tissue
P. aeruginosa 116 (28.5) 121 (45.7) 106 (21.9) 48 (36.9)
E. coli 32 (7.9) 17 (6.4) 162 (33.5) 23 (17.7)
K. pneumoniae 70 (17.2) 32 (12.0) 90 (18.6) 15 (11.5)
A. baumannii 68 (16.7) 24 (9.1) 23 (4.8) 10 (7.7)
E. cloacae 33 (8.1) 12 (4.5) 40 (8.3) 7 (5.4)
P. mirabilis 12 (2.9) 4 (1.5) 24 (5.0) 4 (3.1)
Others 76 (18.7) 55 (20.8) 38 (7.9) 23 (17.7)
Total 407 (100) 265 (100) 483 (100) 130 (100)
www.bjid.com.br
220 The MYSTIC Program Brazil 2003 BJID 2005; 9 (June)
% µ
µg/mL
Species/antimicrobial S I R MIC50 MIC90
P. aeruginosa (n = 470)
Cefepime 58.3 11.7 30 6 >256
Ceftazidime 55.8 5.5 36 4 >256
Imipenem 58.7 4.7 36.6 2 >32
Meropenem 64 2.1 33.9 1 >32
Piperacillin/tazobactam 63.8 0 36.2 24 >256
Ciprofloxacin 49.6 2.1 48.3 1.5 >32
Gentamicin 53.2 2.5 44.3 4 >256
Tobramycin 54 4 42 1.5 >256
Amikacin 63.4 2.8 33.8 4 >256
E. coli (n = 288)
Cefepime 85.4 0 14.6 0.032 6
Ceftazidime 85.4 0 14.6 0.19 3
Cefotaxime 85.4 0 14.6 0.064 64
Imipenem 100 0 0 0.19 0.25
Meropenem 100 0 0 0.016 0.032
Piperacillin/tazobactam 98.6 1 0.4 2 4
Ciprofloxacin 76.7 2.1 21.2 0.006 >32
Gentamicin 88.2 2.8 9 0.5 8
Tobramycin 88.9 3.5 7.6 0.75 8
Amikacin 98.6 0 1.4 1.5 3
K. pneumoniae (n = 262)
Cefepime 48.1 0 51.9 1 48
Ceftazidime 48.1 0 51.9 1 48
Cefotaxime 48.1 0 51.9 4 >256
Imipenem 99.2 0.8 0 0.19 0.25
Meropenem 99.2 0.4 0.4 0.032 0.094
Piperacillin/tazobactam 87 5.7 7.3 4 32
Ciprofloxacin 64.1 5.7 30.2 0.125 >32
Gentamicin 52.3 13.4 34.3 2 128
Tobramycin 53 11.5 35.5 3 48
Amikacin 81.7 11.8 6.5 2 32
A. baumannii (n = 137)
Cefepime 33.6 19.7 46.7 24 >256
Ceftazidime 31.4 9.5 59.1 64 >256
Imipenem 97.1 0 2.9 0.75 2
Meropenem 97.1 0 2.9 1 2
Piperacillin/tazobactam 32.1 12.4 55.5 256 >256
Ciprofloxacin 34.3 0 65.7 >32 >32
Gentamicin 53.3 19 27.7 4 >256
Tobramycin 73 9.5 17.5 1.5 256
Amikacin 36.5 5.8 57.7 128 >256
www.bjid.com.br
BJID 2005; 9 (June) The MYSTIC Program Brazil 2003 221
Table 5. Susceptibility (%) of P. aeruginosa isolates per center – MYSTIC Program Brazil 2003
% Susceptible
Center n CEP CAZ IMP MEM PTZ CIP GM TB AK
1 40 12.5 12.5 10 12.5 35 10 12.5 10 32.5
2 17 94.1 94.1 88.2 94.1 94.1 88.2 88.2 88.2 88.2
4 21 66.7 66.7 61.9 66.7 66.7 61.9 57.1 61.9 61.9
5 26 50 50 19.2 42.3 46.2 19.2 23.1 23.1 57.7
6 19 52.6 47.4 52.6 52.6 47.4 31.6 57.9 52.6 52.6
7 23 56.5 39.1 69.6 69.6 47.8 56.5 56.5 56.5 60.9
8 28 50 60.7 53.5 57.1 60.7 46.4 67.9 78.6 78.6
9 25 68 68 76 76 64 60 64 60 64
10 10 100 80 100 100 90 70 70 70 90
11 16 87.5 87.5 93.8 100 100 93.8 81.3 87.5 100
12 40 60 60 60 70 65 45 57.5 50 60
13 63 69.8 71.4 71.4 76.2 73 69.8 68.3 71.4 73
14 22 31.8 31.8 50 50 54.5 22.7 22.7 22.7 36.4
15 7* 28.6 14.3 28.6 28.6 42.9 14.3 14.3 28.6 28.6
16 43 62.8 55.8 58.1 62.8 65.1 53.5 53.5 55.8 72.1
17 14 57.1 57.1 71.4 71.4 78.6 71.4 71.4 71.4 71.4
18 15 40 100 93.3 93.3 60 33.3 40 40 53.3
19 17 52.9 52.9 47.1 52.9 70.6 47.1 47.1 47.1 64.7
20 5* 60 40 40 40 40 0 0 0 0
21 19 94.7 94.7 68.4 89.5 89.5 68.4 73.7 78.9 78.9
Total 470 58.3 58.5 58.7 64 63.8 49.6 53.2 54 63.4
CEP Cefepime IMP Imipenem PTZ Piperacillin/tazobactam
CAZ Ceftazidime MEM Meropenem CIP Ciprofloxacin
GM Getamicin TB Tobramycin AK Amikacin.
meropenem) gave 100% susceptibility rates against aerogenes (n=23) isolates, 7 (30%) were
isolates of Citrobacter spp., Enterobacter spp., consistent with a phenotype of AmpC
Serratia spp., and Providencia spp. Among hyperproduction, and 3 (13%) of those were also
Citrobacter freundii isolates, 52% (11/21) suggestive of ESBL production. Among Serratia
presented resistance to extended-spectrum marcescens (n=38) isolates, 12 (32%) were
cephalosporins, with a phenotype suggestive of a consistent with a phenotype of AmpC
chromosomal AmpC hyperproducer. Among E. hyperproduction, and 4 (11%) of those were also
cloacae (n=110) isolates, 42 (38%) presented suggestive of ESBL production. Among Proteus
resistance to extended-spectrum cephalosporins, mirabilis (n=51) and Morganella morganii
with a phenotype suggestive of chromosomal (n=12) isolates, 11 (22%) and 3 (25%),
AmpC hyperproducer. Additionally, 27 of those 42 respectively, presented phenotypes suggestive of
presented resistance to cefepime, also suggesting ESBL production, conferring resistance to third and
ESBL production. Among Enterobacter fourth generation cephalosporins.
www.bjid.com.br
222 The MYSTIC Program Brazil 2003 BJID 2005; 9 (June)
Discussion catheter was expected, since our study did not aim at
establishing the prevalence of nosocomial infections.
The MYSTIC Program is a large-scale surveillance But rather, the study aimed at isolating clinically-
program of nosocomial bacterial isolates with significant bacteria causative of the infectious
associated information on their MICs. Our data was processes.
collected from patients hospitalized in 20 hospitals The susceptibility patterns detected by the MYSTIC
located in eight Brazilian states during the 2003 edition Program 2003, particularly for meropenem against P.
of the program. The program’s main objective was to aeruginosa and A. baumannii, in these 20 Brazilian
evaluate the susceptibility pattern of Gram-negative centers demonstrated resistance rates somewhat higher
bacilli isolated from patients with nosocomial infections. than the ones determined by other studies
This is in accordance with the fundamentals of other [4,6,7,12,20,21], although lower than the ones
microbiological surveillance studies, since these studies detected in the 2002 program. The present MYSTIC
aim to identify regional patterns of resistance in specific edition in Brazil showed 36% resistance rate to
settings. Surveillance programs also play a role as major meropenem in P. aeruginosa isolates, while the
contributors to guiding empirical antimicrobial therapy previous 2002 edition showed 40.2% (in press), and
[8,9]. However, these programs are limited in their the others showed resistance rates to carbapenems in
ability to answer all relevant clinical and microbiological P. aeruginosa and A. baumannii ranging between 18-
outcome issues for all world regions, thus reinforcing 21% and 14-15%, respectively [18,19]. Possible
the need for regional data. reasons for the higher resistance patterns observed
Pseudomonas aeruginosa was the most frequently during 2002 and 2003, when compared to other
submitted isolate, accounting for 30.3% of all isolates, editions and to other surveillances, could be based on
followed by E. coli (18.6%), K. pneumoniae (16.9%), the program’s selection of carbapenem user hospital
A. baumannii (8.8%), and E. cloacae (7.1%). The units and of specialized centers, particularly with an
frequency of P. aeruginosa and A. baumannii has increased number of intensive care units during 2003.
risen significantly, when compared to the first MYSTIC Other possible reasons could also be that the centers,
edition in Brazil [18], but it has remained rather constant although scattered around the country, were all major
since the previous editions in 2001 [19] and 2002 (in reference hospitals with specialized units. This may
press). Even with the increase in participating centers reflect a specific influence of demographic
as compared to the 2001 and 2002 editions, the characteristics of these units in the high resistance rates
frequency of P. aeruginosa isolates has remained obtained. Furthermore, clonal spread among P.
constant at around 30%. This may be due to the aeruginosa and A. baumannii was confirmed during
characteristics of the participating centers, which may the 2002 program edition, with documented clonal
be similar, and to the exclusive isolation of Gram- spread within the same centers and among different
negative bacteria during all three years. It should also centers, even in different city locations [22,23].
be noted that, similar to the previous edition, at least Pseudomonas aeruginosa (n=470) isolates
57.4% of samples in the present edition were from presented high resistance rates against all antimicrobials,
clinically significant sources and definitely related to the with descending order of susceptibility rates of 64% to
infectious process (blood, catheter and urinary tract). meropenem (MIC50 1µg/mL), 63.8% to piperacillin/
However, 17.1% of samples were from the respiratory tazobactam (MIC50 24µg/mL), 63.4% to amikacin
tract, although always considered by investigators as (MIC50 4µg/mL), 58.7% to imipenem (MIC50 2µg/
causative agents of the infectious processes. mL), 58.3% to cefepime (MIC50 6µg/mL), and 55.8%
Nevertheless, one cannot completely rule out the to ceftazidime (MIC50 4µg/mL). Significant differences
contribution of colonizers as part of the total amount of in resistance rates were observed among the centers
isolates. The higher number of isolates from blood/ from different regions, but in our analysis it was not
www.bjid.com.br
BJID 2005; 9 (June) The MYSTIC Program Brazil 2003 223